Tensorflow.js tf.train.sgd() Function

Tensorflow.js is an open-source library that is developed by Google for running machine learning models as well as deep learning neural networks in the browser or node environment.

The .train.sgd() function is used to build a tf.SGDOptimizer which utilizes stochastic gradient descent.

Syntax:

tf.train.sgd(learningRate)

Parameters:

• learningRate: The stated learning rate which is used in favor of the SGD algorithm. It is of type number.

Return Value: It returns tf.SGDOptimizer.

Example 1:

Javascript

// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"

// Declaring learningRate
const learning_rate = 0.02;

// Calling train.sgd() method
const res = tf.train.sgd(learning_rate);

// Printing output
console.log(res);

Output:

{
  "learningRate": 0.02,
  "c": {
    "kept": true,
    "isDisposedInternal": false,
    "shape": [],
    "dtype": "float32",
    "size": 1,
    "strides": [],
    "dataId": {
      "id": 1298
    },
    "id": 1192,
    "rankType": "0",
    "scopeId": 1251
  }
}

Example 2:

Javascript

// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"

// Calling train.sgd() and sin() method
console.log(JSON.stringify(tf.train.sgd(tf.sin(45))));

Output:

{"learningRate":{"kept":false,"isDisposedInternal":false,"shape":[],
"dtype":"float32","size":1,"strides":[],"dataId":{"id":1316},"id":1207,
"rankType":"0","scopeId":1269},"c":{"kept":true,"isDisposedInternal":false,
"shape":[],"dtype":"float32","size":1,"strides":[],"dataId":{"id":1318},
"id":1208,"rankType":"0","scopeId":1269}}

Reference: https://js.tensorflow.org/api/latest/#train.sgd